Absolute Versus Relative Changes in Cardiac Troponin T: Corresponding Cutoffs Based on Quantile Generalized Additive...

Area of Science:

• Cardiology
• Biomarker Analysis
• Diagnostic Accuracy

Background:

• Myocardial infarction (MI) diagnosis typically relies on relative changes in cardiac troponin.
• Absolute change cutoffs for cardiac troponin offer potentially greater diagnostic sensitivity.
• Establishing precise absolute change thresholds for high-sensitive cardiac troponin T (hs-cTnT) is crucial.

Purpose of the Study:

• To determine the absolute changes in hs-cTnT (absΔhs-cTnT) corresponding to established relative change cutoffs (relΔhs-cTnT).
• To enhance the diagnostic accuracy of myocardial injury/infarction using optimized hs-cTnT measurements.

Main Methods:

• Utilized a quantile generalized additive model (qgam) to fit the relationship between absΔhs-cTnT and relΔhs-cTnT.
• Employed ordered quantile-based normalization (OQN) on plasma Δhs-cTnT data from 9,753 patients (1-6 hour intervals) to mitigate extreme value influence.
• Analyzed a 6-year dataset of patient plasma samples.

Main Results:

• The qgam regression curve demonstrated a nonlinear relationship between absolute and relative hs-cTnT changes.
• MI diagnosis rates were comparable and more reliable when using the qgam model with OQN-transformed data versus untransformed data.
• The optimized qgam approach provided specific absolute Δhs-cTnT values corresponding to relative Δhs-cTnT cutoffs.

Conclusions:

• An optimized qgam-based method effectively accounts for heavy-tailed distributions in hs-cTnT data.
• This approach successfully provides absolute Δhs-cTnT values for corresponding relative Δhs-cTnT cutoffs.
• The findings support the use of absolute hs-cTnT changes for more sensitive and reliable MI diagnosis.